Around 2017, we were contacted by the cryo-EM facility at Arizona State University to discuss our use of a recently installed Titan Krios. The microscope had been used extensively, but there were no users focusing on tomography, an approach we had proposed when requesting funds for the Krios. After a long meeting discussing how to image Geobacter sulfurreducens cells, we realized this effort was going to be challenging. We were interested in imaging conductive nanowires used by G. sulfurreducens to connect to an anode, an easy task for the cryo-EM. But the ideal tomography sample is a single cell. As G. sulfurreducens grows in biofilms at an anode, grabbing a single cell is nearly impossible. Breaking cells apart would most likely result in breaking nanowires as well. So, the next year was spent on developing a specific protocol for imaging G. sulfurreducens: using a cryo-EM grid as the anode for growth. Cryo-EM grids are 3 mm in diameter. Our first task was to design a grid holder that would electrically connect several grids to a current collector and a wire coming out of our electrochemical cell. We tested various factors before settling on gold grids placed in our electrochemical cells for 18-24 hrs. This was enough time to observe singular active cells throughout the grid. With the goal of imaging G. sulfurreducens nanowires, we now had a protocol that would reliably place cells at the grid and nanowires could be imaged connecting to the anodic cryo-EM grid.
Figure 1. Top – schematic of the grid holder design to grow G. sulfurreducens in cryo-EM grids. Bottom –cryo-EM grids imaged through confocal microscope after 72 hrs showing initial developments of a biofilm, especially at the gold surface that can be seen as dark bars in the transmitted light image (right). Cells images are on top of the grid holes that have a thin layer of holey carbon.
Our paper is not about microbial nanowires, but about intracytoplasmic membranes (ICMs) in G. sulfurreducens. We serendipitously came into this discovery. Most cells we started imaging with Titan Krios had this internal lipidic-like structure inside the cytoplasm. It became difficult to focus on nanowires! Our first questions were: what are these? and why haven’t they been seen before? Identifying these structures as ICMs was not that simple. To our knowledge, these are the first cryo-tomograms of ICMs, in general; most ICMs have been imaged through various approaches in resin-embedded TEM efforts. As the reader can observe, the structure of the ICMs in cryo-embedded TEM is quite different from our cry-EM tomograms. It was not until we imaged through TEM that we could see clear resemblances to ICMs reported in ammonia oxidizers and/or methane oxidizers. Once identified, we could follow similar approaches to characterize the ICM, such as the lipid staining that has been reported before.
One question remained, why are we now observing these ICMs, even in TEM slices, when G. sulfurreducens is the most studied electrogenic microorganism? There are dozens of papers showing some sort of high-resolution imaging of G. sulfurreducens and none had reported or showed evidence of these structures. Trying to answer this question led to an important hypothesis in our research: G. sulfurreducens produces ICMs at low potentials. Interestingly, our team is one of very few that commonly grow G. sulfurreducens at anode potentials below 0 V vs. standard hydrogen electrode (SHE). In fact, our default growth potential is -0.08 V vs. SHE. The reason for this potential stems back to our core in engineering applications of microbial electrochemistry, where G. sulfurreducens is typically operated at potentials below -0.1 V. The limiting potentials are not optimal for growth, which is why most groups use 0.2 V vs. SHE or higher. It turns out that the production of ICM is associated with slow respiratory proteins in other microorganisms, such as ammonia and methane monooxygenases known to be present in ICMs. At high potentials and with plenty of energy, G. sulfurreducens should not have a limiting respiratory protein and has no need for an ICM. But once you limit energy by providing a low potential, respiratory proteins at the inner membrane can slow down, leading to a need for more proteins and thus more inner membrane. The ICM is then developed.
Figure 2. Additional cryo-tomogram and resulting 3D models of G. sulfurreducens showing an ICM near the tip of the cell and putative nanowires extending from the tip.
We are still far from understanding the use of ICMs in G. sulfurreducens. If indeed these are respiratory centers, as our hypothesis suggest, how are electrons transported outside the cell for extracellular respiration? Are ICMs only associated with low-potential pathways? And is there a relationship between ICM production and biofilm organization? Overall, the discovery of ICMs allows a better understanding of how G. sulfurreducens can thrive with limited available energy in electrochemical cells, paving the way for better prediction of their behavior in applied systems.